PRESSURE-TEMPERATURE DETERMINATIONS FROM INTEGRATED ELASTIC AND TRACE ELEMENT THERMOBAROMETRY: A CASE STUDY FROM THE (U)HP TERRANE OF PAPUA NEW GUINEA

Elastic thermobarometry may be used to constrain the P-T conditions at the time of inclusion entrapment and does not require assumptions of mineral assemblage equilibrium. Despite its promise, application of elastic thermobarometry to natural rocks has been limited, possibly because temperature estimates are difficult to relate to the timing of inclusion entrapment. We propose a method of integrating multiple in-situ thermobarometric techniques on quartz inclusions in garnet to determine P-T conditions during inclusion entrapment. Raman spectroscopy was used to constrain remnant pressures from elastically isolated quartz inclusions in garnet from a quartzofeldspathic gneiss that contains Pliocene age eclogite boudins from Goodenough Island, Papua New Guinea. Inclusion remnant pressures were used in an elastic model to constrain the possible P-T conditions of entrapment (i.e. an isomeke). Inclusions have measured remnant pressures of 2.1-3.0 kbar. Elastic modeling yields a group of isomekes interpreted as representing garnet growth at similar metamorphic conditions. Ti-in-quartz thermobarometry on 20-40 μm quartz inclusions exposed at the surface of garnets yielded average [Ti] of 18.1 ± 1.5 ppm. The integrated results yield a P-T estimate of 9.9 ± 1.25 kbar and 590 ± 40^oC. We evaluated the possible effects of Ti diffusion and the relevancy of the Ti-in-quartz thermobarometric constraint to the elastic models. Since the characteristic diffusion length scale for Ti in quartz is 10 μm at 600^oC for a 1 Ma duration, diffusive modification of Ti should be discernible at the quartz inclusion rims. No [Ti] gradients were observed, indicating that diffusive modification did not affect the [Ti] in quartz inclusions. Results illustrate that integration of elastic thermobarometry isomekes and trace element thermobarometry isopleths on the same host-inclusion system can be used to constrain the metamorphic conditions of inclusion entrapment. The P-T estimates from elastic and trace element thermobarometry are interpreted to indicate that garnet grew during exhumation of the quartzofeldspathic host gneiss, and not at peak (U)HP conditions. Results are consistent with numerical models of (U)HP rock exhumation paths resulting from retreat of an oceanic upper plate away from the subduction channel.